Most evidence indicates that G protein-coupled receptors (GPCRs) can form homo and heteromers. Homodimers seem to be a predominant species with potential dynamic formation of higher-order oligomers (1,2). The pentameric structure constituted by one GPCR homodimer and one heterotrimeric G protein may provide a main functional unit at the plasma membrane and oligomeric entities can be viewed as multiples of dimers (1,2). Our recent studies indicate that GPCR heteromers are often constituted by heteromers of two homodimers, i.e. heterotetramers (1,2). Allosteric mechanisms determine a multiplicity of possible unique pharmacological properties of GPCR homomers and heteromers. Some general mechanisms seem to apply, particularly at the level of ligand-binding properties, but also unique properties for GPCR heteromers emerge in relation to different intrinsic efficacy of ligands for different signaling pathways (functional selectivity) (1,2). We are interested in GPCR heteromers localized in the brain and, particularly, in brain circuits involved in substance use disorders (3). Most studies on GPCR oligomerization are performed in artificial cell systems. The main hurdle in the field is demonstrating their existence and functional significance in situ, in native tissues, and, even more challenging, in vivo, in the experimental animal and in humans. Our strategy is to develop tools to disrupt intermolecular interactions between the receptor units (protomers) in the GPCR heteromer and find allosteric properties that are dependent on those intermolecular interactions. The presence of those allosteric modulations in brain tissue demonstrates the presence of the GPCR heteromer. In our laboratory we have demonstrated that synthetic peptides with the amino acid sequence of transmembrane domains (TM; fused to an HIV TAT peptide to promote integration in the plasma membrane) are powerful tools to disrupt GPCR oligomerization (4,5). Furthermore, we have developed a new microdialysis probe that allows the direct infusion of TM peptides in the freely moving animal, allowing a more direct demonstration of GPCR heteromers in vivo (5). Previous studies from our group demonstrated the ability of adenosine A2A receptor (A2AR) and dopamine D2 receptor (D2R) to form heteromers and strongly suggested their presence in the striatum (2). We also suggested that an agonist-agonist allosteric interaction in the A2AR-D2R heteromer, by which A2AR agonists decrease the affinity of D2R agonists, provides a key mechanism involved in the psychostimulant effects of caffeine and a rationale for the use of A2AR antagonists in Parkinson's disease (2). Using disrupting TM peptides in artificial cell systems and striatal tissue, we have been able to demonstrate a heterotetrameric quaternary structure of the A2AR-D2R (4). We have then revisited the allosteric interactions within the A2AR-D2R heteromer within the new frame of GPCR heterotetramers. Our results allowed us to demonstrate the existence of multiple allosteric and reciprocal interactions between A2AR and D2R ligands, which not only involve modulation of ligand affinity but also differential modulation of the intrinsic efficacy for G protein-dependent and independent signaling (functional selectivity of allosteric modulation within GPCR heteromers) (6). These properties make the A2AR-D2R heterotetramer a cellular device that integrates signals from the extracellular and extracellular compartments (dopamine, adenosine and calcium) to produce a specific functional response (6). From the results of these and previous experiments with dopamine D1-D3 receptor heteromers we now postulate that the canonical interaction between activated Gs and Gi proteins at adenylyl cyclase signaling is in fact a biochemical property of GPCR heteromer (1). The heterotetramer model also allowed understanding unexpected findings we observed both in artificial cell systems and in striatal tissue: the ability of A2AR antagonists, including caffeine, to also exert the same allosteric modulations of D2R ligands than A2AR agonists, while A2AR agonists and antagonists counteract each others effects (4). The model implies that, in the A2AR-D2R heteromer, hybrid occupancy of the A2AR dimer by an agonist and an antagonist does not modulate D2R agonist binding to the D2R dimer. The model accurately predicted that high concentrations of A2AR antagonists would behave as A2AR agonists and decrease D2R function in the brain (4). Unexpectedly also, we found that both and A2AR agonist and caffeine significantly reduce D2R antagonist binding in membrane preparations from sheep and human striatum and from mammalian cells transfected with A2AR and D2R. Furthermore, low concentrations of caffeine antagonized the effect of the A2AR agonist, while high concentrations were also associated with a significant decrease in D2R antagonist binding (4). TM peptides shown to destabilize the quaternary structure of the A2AR-D2R heteromer in transfected cells selectively destabilized A2AR-D2R heteromerization in sheep striatal slices, as analyzed by a proximity ligation assay and caffeine-mediated decrease D2R antagonist binding in sheep striatal membrane preparations (4). Our results indicate that, in the A2AR-D2R heterotetramer, any orthosteric A2AR ligand, agonist or antagonist, can decrease the affinity and intrinsic efficacy of any orthosteric D2R ligand, agonist or antagonist, and that A2AR agonists and antagonists counteract their allosteric effects on D2R ligands. This could explain our results originated from a PET study in humans, where acute administration of caffeine produced a significant increase in the binding of the D2R antagonist 11Craclopride in putamen and ventral striatum, when compared to placebo (7). These results call for the need of monitoring caffeine intake when evaluating the effect of D2R ligands in humans, when used as therapeutic agents in neuropsychiatric disorders or as probes in imaging studies. Release of the neuropeptides corticotropin-releasing factor (CRF) and orexin-A in the ventral tegmental area (VTA) play an important role in stress-induced cocaine-seeking behavior. Using the same in vitro techniques described above and the modified microdialysis-infusion probe for in vivo delivery of CRF, orexin-A and disrupting TM peptides, we provided in vivo evidence for pharmacologically significant interactions between CRF and orexin-A that depend on oligomerization of CRF1 and orexin OX1 receptors (5). CRF1R-OX1R heteromers were the conduits of a negative crosstalk between orexin-A and CRF as demonstrated in transfected cells and in the VTA, where they significantly modulate dendritic dopamine release. The cocaine target sigma &#963;1 receptor (&#963;1R) also associated with the CRF1R-OX1R heteromer and cocaine binding to the &#963;1R-CRF1R-OX1R complex promoted a long-term disruption of the orexin-A-CRF negative crosstalk (5). Through this mechanism cocaine sensitizes VTA cells to the excitatory effects of both CRF and orexin-A, thus providing a mechanism by which stress induces cocaine seeking. Therefore CRF1R-OX1R heteromers in the VTA can constitute new targets for drug development for cocaine use disorder.